The 10-Minute Claude Habit That Saves Me Hours Every Week

The single highest-leverage Claude habit I have is not a prompting trick. It is ten minutes spent at the start of a task writing down what "done" actually means, before asking Claude to do anything. Almost everyone skips this because it feels like overhead. It is the opposite of overhead — it is the thing that prevents three rounds of "not quite what I meant."

What the ten minutes actually looks like

Write, in plain language: what the end state looks like, what must not break, and how you will check that it worked. That is it. No formal spec, no template — just three answers, written down before the first prompt goes out.

Why "done" is almost always underspecified in your head

You know what you mean by "add a login page." Claude does not know whether that includes password reset, whether it needs to match an existing design system, or whether "login" means email-and-password or something else entirely. The ambiguity was always there — writing it down just makes you notice it before it costs you a wasted turn.

This habit pays for itself even on tasks that feel too small for it

The instinct is to skip this step on "quick" tasks, which is exactly when it is skipped most and needed least obviously — a five-minute task with a hidden ambiguity becomes a twenty-minute back-and-forth. The habit costs a flat ten minutes; the alternative costs a variable, usually larger amount, hidden as several rounds of correction.

The version that scales to bigger work

For anything more involved than a single small task, this same habit is what a proper plan is for — writing down the approach and having it reviewed, or interrogated, before a single line of implementation happens. The ten-minute version is the minimum viable form of the same idea.

Ten minutes of writing down what done means, every single time, beats hours of cleaning up a task that was never actually specified — for the person asking as much as for whatever is doing the work.